Building KubeOne & Kontemplate the simpler yet powerful Helm alternative?

München Kubernetes/Cloud-Native Meetup
München Kubernetes/Cloud-Native Meetup
Öffentliche Gruppe

eGym GmbH

Einsteinstraße 172 · München

Wie du uns findest

The office, located at the Blue Tower is easily accessible following google maps directions. Go to the 2nd floor where the room is.

Bild des Veranstaltungsortes

Details

Hi Kubernauts,

Our next Meetup in München will take place on 13th August @eGym.

Agenda:
6:30pm - Welcome
7:00 pm - Alexander Sowitzki, Loodse
7:30 pm - Max Körbächer, Storm Reply
8:00 pm - Food, drinks & networking

******************************************************************
Alexander Sowitzki, Loodse

"Building KubeOne: yet another Kubernetes cluster management tool"

At Loodse, we have recently announced a new open source Kubernetes cluster management tool called KubeOne. We had a vision of replacing the various imperative tools we used before, with a new single tool that would allow us to manage our clusters in a cloud native way. We wanted to be able to declare clusters in form of a manifest, just like our workload, and have a tool that would bring us a ready-to-use cluster.

But how to go about building such a tool? There are many existing tools as well as solutions and that was precisely out problem. It's hard to pick the right one that would satisfy all our requirements. That’s why we decided to build a new tool atop of solutions like kubeadm and Cluster-API. This talk is about building a new cluster management tool and the challenges this presented us with.

******************************************************************
Max Körbächer, Storm Reply

"Kontemplate the simpler yet powerful Helm alternative?"

Doing Kubernetes right means mostly to have a bunch of work to do. You’re not only doing a deployment but you also will need a service account, RBAC, expose your pod, network policies and so on. That's where Helm takes a lot of work from you, but also bring his own style of complexity. Kontemplate is a lightweight alternative supporting multiple use cases from variable interpolation to complex template logic, and we will have a closer look at it and how we use it in production.

******************************************************************

See you there.